arXiv:hep-th/0611183v3 28 Jan 2007 SU-ITP-2006-31 November 16, 2006 O’KKLT Renata Kallosh and Andrei Linde Department of Physics, Stanford University, Stanford, CA 94305 Abstract We propose to combine the quantum corrected O’Raifeartaigh model, which has a dS minimum near the origin of the moduli space, with the KKLT model with an AdS minimum. The combined effective N=1 supergravity model, which we call O’KKLT, has a dS minimum with all moduli stabilized. Gravitino in the O’KKLT model tends to be light in the regime of validity of our approximations. We show how one can construct models with a light gravitino and a high barrier protecting vacuum stability during the cosmological evolution.
13
Embed
Renata Kallosh and Andrei Linde arXiv:hep-th/0611183v3 28 ... · 1− cλ2 16π 2log(1+ λ2 S¯ m) for λ2SS¯ ≪ 1. Here Λ2 = 16π2m2 cλ4, c = O(1), and we assume that λ2 16π2
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
arX
iv:h
ep-t
h/06
1118
3v3
28
Jan
2007
SU-ITP-2006-31November 16, 2006
O’KKLT
Renata Kallosh and Andrei Linde
Department of Physics, Stanford University, Stanford, CA 94305
Abstract
We propose to combine the quantum corrected O’Raifeartaigh model, which has adS minimum near the origin of the moduli space, with the KKLT model with an AdSminimum. The combined effective N=1 supergravity model, which we call O’KKLT, hasa dS minimum with all moduli stabilized. Gravitino in the O’KKLT model tends to belight in the regime of validity of our approximations. We show how one can constructmodels with a light gravitino and a high barrier protecting vacuum stability during thecosmological evolution.
The 4th term in (2.10) consists of 2 terms. The first one does not depend on α, whereas the
second one depends on cos(aα), but it is significantly smaller than the KKLT α-dependent term
shown in eq. (2.11). There are no other terms linear in y in the potential. All other terms
are either quadratic or higher power in y2. Therefore the linearized equation for y and exact
equation for α extremizing the potential (at fixed values of σ and x) take the following form
∂V
∂α= c1 sin(aα) + c2y cos(aα) ,
∂V
∂y= c3 sin(aα) + c4y . (2.13)
This shows that the point α = y = 0 remains the extremum of the potential after the uplifting.
One can also show that it remains a minimum.
The potential at α = y = 0 is significantly simplified. One can now proceed with evaluation
of the change of its minimum with respect to σ and x. Since the KKLT potential is uplifted
at least via the second term in (2.10) we know that σ changes a bit, as in the original KKLT
model. This shift is small, exactly as in the original KKLT model where the uplifting occurs
due to the D3 brane.
The situation with the field x, the real part of S, is the following. One can expand the
potential in powers of x. The coefficients in such expansion are complicated functions of σ.
However, in the first approximation, we can calculate the values of the linear and quadratic
terms in x at σ = σ0 where σ0 is the value of σ at the AdS minimum before the uplifting. At
this point we find that
V ≈ c0(σ0) + c1(σ0)x− c2(σ0)
2x2 + ... (2.14)
According to (2.6), (2.10), the O’Raifeartaigh model uplifts the AdS minimum with the depth
|VAdS| ≈ 3m23/2 [22] by µ4
(2σ0)3, so that µ4
(2σ0)3+ VAdS = c0(σ0) ∼ 10−120 ≈ 0. This implies that
m23/2 ≈
µ4
3(2σ0)3. (2.15)
The value of the field x in the uplifted minimum is determined by the condition V ′ = 0,
which gives
x0 ≈c1(σ0)
c2(σ0)=
√3Λ2
6− Λ2≈
√3
6Λ2 . (2.16)
In the derivation of this formula we used the total expression for the combined potential from
Mathematica and the values of the σ0 from [10] in the form aAe−aσ0 = m3/2
√18σ0.
This result has an interesting and instructive interpretation in terms of a simpler model
recently studied by Kitano [8]. He studied the supergravity model with
WK = −µ2S + C, KK = SS − (SS)2
Λ2. (2.17)
The difference with what we call quantum corrected O’Raifeartaigh model is the presence of
the constant term C in the superpotential. Without the term (SS)2
Λ2 in the Kahler potential this
4
would be the supergravity Polonyi model [24], which is known to have a Minkowski vacuum at
the fine-tuned Planckian value of the field S. Thus, the model (2.17) is a hybrid of the Polonyi
model and the quantum corrected O’Raifeartaigh model. It was shown by Kitano that one can
fine-tune the constant C to get the Minkowski vacuum, as in the Polonyi model. However, in
this hybrid case the minimum of the potential appears at not at S = O(1), as in the Polonyi
model, but at S ≈√36Λ2 ≪ 1 [8]. The total potential based on (2.17) at small x and small Λ2
can be represented, using [23], in a compact form:
VK ≈ µ4[(1− 3C2)− 4Cx+ 2(2
Λ2− C2)x2 + ... (2.18)
The field x is stabilized at x0 ≈ CΛ2
2−C2Λ2 . If C2 is tuned to Minkowski vacuum, C2 = 1/3, one
finds for small Λ2 that x0 ≈√3
6−Λ2Λ2 ≈
√36Λ2, precisely as in the O’KKLT model.
This result is pretty general; in particular, it is valid for the generalized KKLT models to
be discussed in the next section. The meaning of this result is that the KKLT model supplies
the superpotential of the quantum corrected O’Raifeartaigh model by the constant term (at
fixed σ = σ0). In terms of Eqs. (2.17), WKKLT (σ0) = C serves for adjusting the height
of the minimum and making it the (nearly) Minkowski one. In other words, fine tuning of
WKKLT (σ0) = C in the O’KKLT model allows us to achieve the cancellation between the
negative energy density in the AdS minimum of the KKLT and the positive energy density in
the dS minimum of the quantum corrected O’Raifeartaigh model.
100120
140
160
180
200
Σ0
0.0001
0.0002
0.0003
0.0004
x0
0.5
1
1.5
2
V100
120140
160
180
200
Σ
Figure 1: The slice of the O’KKLT potential at vanishing axions y and α, multiplied by 1031, for thevalues of the parameters A = 1, a = 0.25, W0 = −10−12, µ2 = 1.66×10−12, L = 10−3. The potentialhas a dS minimum at σ ≈ 123, x ∼ 3 · 10−7. The gravitino mass in this example is m3/2 ∼ 600 GeV.
5
Our result x0 ≈√36Λ2 implies that the consistency condition SS = x2 ≪ m2
λ2 is satisfied for
m . 10−2λ3 . (2.19)
The only parameters which are required for the calculations of the potential on the O’Raifeartaigh
side are µ and Λ, whereas m = Λ√c
4πλ2, and λ is a free parameter, which can be varied in order
to satisfy the consistency conditions. One can easily find many sets of parameters which satisfy
all required conditions. In particular, one may find the theory with the gravitino mass in the
TeV range if one takes µ ∼ 10−6, Λ ∼ 10−3, m = 10−6, λ ∼ 10−1, see Figure 1.
Thus the O’KKLT model can provide a consistent model of the F-term uplifting with the
gravitino mass in the TeV range. On the other hand, if one attempts to describe superheavy
gravitino, then it becomes difficult to satisfy all the consistency conditions, which require that
some of the mass scales must be sufficiently small. In order to describe superheavy gravitino
in such a model one may need to go beyond the simple approximations used to calculate the
scalar potential and the Kahler potential in our scenario.
3 Light gravitino, vacuum stability and the KL model
For many years, we wanted to have models with m3/2 in the TeV range, and the O’KKLT
model is well suited for this purpose. This fact is quite interesting, especially if one compares
our model with the models with the D-term uplifting, where it is very difficult to obtain a light
gravitino [11]. In the O’KKLT model one can obtain a light gravitino, but it may be difficult
to obtain a very heavy one.
This property of our model may be desirable from the point of view of the phenomenological
supergravity, but in the KKLT context it may lead to some cosmological problems [22]. Indeed,
the depth of the AdS vacuum in the KKLT scenario is given by VAdS = −3eK |W |2 = −3m23/2.
Here VAdS is the depth of the AdS minimum prior to the uplifting, and m3/2 is the gravitino
mass after the uplifting. Uplifting creates the barrier separating the KKLT minimum from the
10D Minkowski Dine-Seiberg minimum. The height of the barrier is somewhat smaller than
the depth of the original AdS minimum prior to the uplifting:
Vbarrier ∼ |VAdS| ≈ 3m23/2 . (3.1)
Inflation requires the existence of an additional contribution Vinfl to the scalar potential, but all
such contributions in the effective 4D theory have the following general structure: Vinfl = V (φ)(ρ+ρ)3
,
where φ is the inflaton field; see, for example, the second term in (2.10). Such terms destabilize
the potential for Vinfl ≫ Vbarrier ∼ m23/2, see Fig. 2. Since during inflation one has Vinfl = H2/3,
one finds the constraint on the Hubble constant during the last stage of inflation [22]
H . m3/2 . (3.2)
If gravitino is heavy, m3/2 ∼ 1011 − 1016 GeV, the scale of inflation can be very high, and
there is no destabilization of the volume modulus during inflation. However, if the mass of
6
gravitino is light, e.g. m3/2 ≤ 1 TeV, one would need to consider non-standard low-scale
inflationary models. Such models do exist, but still the constraint H . m3/2 is quite restrictive
and undesirable.
100 150 200 250 Σ
1
2
3
4
V
Figure 2: The lowest curve with dS minimum is the one from the uplifted KKLT model. The second
one describes the inflationary potential with the term Vinfl = V (φ)σ3 added to the KKLT potential. The
top curve shows that when the inflationary potential becomes too large, the barrier disappears, andthe internal space decompactifies. This explains the constraint H . m3/2.
This problem was addressed in our paper [22], which we will call the KL model, to distinguish
it from the simplest KKLT scenario. We used the same Kahler potential as in the KKLT model,
but instead of the superpotential with one exponent we used the racetrack-type superpotential
with two exponents:
K = −3 ln[(ρ+ ρ)] , W = W0 + Ae−aρ +Be−bρ . (3.3)
The potential of this model can describe either one or two AdS vacua, or, with extra fine tuning,
one Minkowski vacuum and one AdS vacuum, both at a finite distance in the moduli space [22].
Here we will be interested in the case of two AdS vacua, such that |VAdS1 | ≪ |VAdS2 |.Indeed, with one exponent in the superpotential as in model (2.7) one cannot simultaneously
solve both equations DρW = 0 and W = 0, which is necessary to get a Minkowski minimum.1
However, with two exponents, as in model (3.3), this is possible. The solution requires the
following relation between the parameters of the superpotential:
−W0 = A
∣
∣
∣
∣
aA
bB
∣
∣
∣
∣
a
b−a
+ A
∣
∣
∣
∣
aA
bB
∣
∣
∣
∣
a
b−a
. (3.4)
The Minkowski minimum occurs at σ = 1a−b
ln∣
∣
aAbB
∣
∣; see Fig. 3.
1There is a no go theorem proposed in [25, 13, 14], which states that it is impossible to have a dS/Minkowskiminimum in the theory with the Kahler potential K = −3 ln[ρ+ ρ], for any choice of the superpotential. Theproof of this theorem is correct for the supersymmetry breaking Minkowski vacua with DρW 6= 0,W 6= 0 [13],but it does not apply to the KL model, where DρW = 0,W = 0 in the Minkowski minimum with unbrokensupersymmetry [22]. It may be useful to remind here that the use of the supergravity function G = K + ln |W |2in studies of vacua with W = 0 should be avoided, as explained in [26], and such vacua should be studiedseparately.
7
60 80 100 120 140 Σ
-1
1
2
3
4V
Figure 3: The KL potential multiplied by 1014, with the parameters A = 1, B = −1.03, a =2π/100, b = 2π/99, W0 = −2 × 10−4. The first minimum, corresponding to the supersymmetricMinkowski vacuum, stabilizes the volume at σ ≈ 62. If one slightly changes the parameters (e.g. takesB = 1.032), this minimum shifts down, and becomes a very shallow AdS minimum, see the thin blackline in Fig. 4.
It has been observed in [27] that it is very easy (for example, by changing slightly the
parameter B in the superpotential) to get models with extremely light gravitino and the AdS
minimum replacing the exact Minkowski one with very small value of |VAdS| but large barrier
separating this AdS from the next one and from the Minkowski vacuum at infinity. To make
this model viable we need to uplift it to dS minimum by one of the available mechanisms. The
D3-brane uplifting is possible. The D-term uplifting in a model with two exponents was not
performed so far. It may require some effort to make it consistent with the gauge invariance of
the superpotential. However, the O’Raifeartaigh uplifting works well for this model.
The supergravity potential is based on
W = W0 + Ae−aρ +Be−bρ − µ2S, K = −3 ln(ρ+ ρ) + SS − (SS)2
Λ2. (3.5)
The complete potential V (σ, α, x, y) as a function of 4 scalars, ρ = σ + iα and S = x + iy, is
again easily computable in Mathematica, using [23], and, as before, we are interested only in
the region of small SS. Investigation of this regime for the KL model practically coincides with
the investigation for the O’KKLT model performed in the previous section. The axion fields
vanish before and after the unification of the O’Raifeartaigh model with the KL model. The
field S after the uplifting takes exactly the same value as in the KKLT scenario: |S| =√36Λ2.
The influence of the O’Raifeartaigh uplifting of a shallow AdS minimum on the position of
this minimum in this scenario is even smaller than the corresponding influence in the O’KKLT
model. Indeed, since the required magnitude of the uplifting in this scenario is much smaller
than the height of the barrier, all parameters of the O’Raifeartaigh model can be taken many
orders of magnitude smaller than the parameters of the KL model.
An example of the KL potential before and after the uplifting (thin and thick lines) is shown
in Fig. 4. The depth of the shallow AdS minimum prior to the uplifting (thin line) corresponds
8
60 80 100 120 140 Σ
-0.5
0.5
1
1.5V
Figure 4: The thin black line shows the potential in the KL model, multiplied by 1014, for the valuesof the parameters A = 1, B = −1.032, a = 2π/100, b = 2π/99, W0 = −2 × 10−4. The shallow AdSminimum (almost Minkowski) stabilizes the volume at σ ≈ 67. The thick blue line shows the potentialafter the O’Raifeartaigh uplifting with µ2 = 0.66 × 10−4, L = 10−3. The AdS minimum after theuplifting becomes a (nearly Minkowski) dS minimum.
to 3m23/2. This depth, and the required magnitude of uplifting, controlled by the parameter µ2,
can be made arbitrarily small by a slight change of the parameter B, which practically does
not affect the height of the barrier. Therefore in this scenario the barrier can be many orders
higher than m23/2. In this figure we have made a relatively large modification of B, which leads
to the large gravitino mass, but we did it only for the purpose of making the modification of
the potential visible.
The main new feature of the KL model as compared with the O’KKLT model is that one
can fine-tune the gravitino mass squared to be extremely small as compared to the height of
the barrier. This allows inflation with H ≫ m3/2 [22].
4 Discussion
The existence of a tiny positive cosmological constant makes it quite important to stabilize all
moduli in a dS state with a positive vacuum energy. In this note we proposed to combine two
simple models to achieve this purpose. The KKLT model originating from string theory brings
in the idea of ∼ 10500 various vacua, mostly supersymmetric AdS vacua with negative energy.
Stabilization of closed string theory moduli is due to non-perturbative effects, like gaugino
condensation or string instantons. The second ingredient of the proposed unified model is a
generic model of dynamical supersymmetry breaking where non-perturbative corrections play
an important role in stabilization of moduli (these moduli in string theory may come from the
open string sector). A typical representative of such models is the O’Raifeartaigh model with
the Coleman-Weinberg quantum corrections. Moduli stabilization in this model occurs near
the origin of the moduli space and results in the existence of a dS vacuum with positive energy.
When these two models are unified, they affect each other in a minor way in all respects but
9
one: the negative AdS energy of the KKLT model is nearly compensated by the positive dS
energy of the O’Raifeartaigh model, which leads to a nearly Minkowski space which we observe.
In the paper we present the detailed description of the unification of KKLT model with
quantum corrected O’Raifeartaigh model, which we called O’KKLT. The effect which each of
these two models has on the other one is computable. In particular, one can find a class of the
O’KKLT models with light gravitinos, and achieve vacuum stability during the cosmological
evolution.
Acknowledgments
It is a pleasure to thank S. Dimopoulos, M. Dine, B. Freivogel, R. Kitano, L. Susskind, J. Wacker
and E. Witten for stimulating conversations. This work was supported by NSF grant PHY-
0244728.
References
[1] L. O’Raifeartaigh, “Spontaneous Symmetry Breaking For Chiral Scalar Superfields,” Nucl.
Phys. B 96, 331 (1975).
[2] M. Huq, “On Spontaneous Breakdown Of Fermion Number Conservation And Supersym-
metry,” Phys. Rev. D 14, 3548 (1976).
[3] E. Witten, “Mass Hierarchies In Supersymmetric Theories,” Phys. Lett. B 105, 267 (1981).
[4] E. Poppitz and S. P. Trivedi, “Dynamical supersymmetry breaking,” Ann. Rev. Nucl. Part.
Sci. 48, 307 (1998) [arXiv:hep-th/9803107].
[5] Z. Chacko, M. A. Luty and E. Ponton, “Calculable dynamical supersymmetry breaking on